wenc

39 ms·

601.
by wenc 3y ago
I think what may have slowed you down is the female TSA agent, not Precheck in general. That’s unfortunate. Also business traveler tip for future: all heavy and light tops (light jackets, suit jacket, winter coats, sweatshirts, hoodies, etc
602.
by wenc 3y ago
I travel a fair bit and have TSA Precheck. It’s very far from spotty. It’s not so much about what you do in line as it is about the line — it’s way way shorter than the normal line. I get through TSA Precheck lines in under 10 minutes where
603.
by wenc 3y ago
Unless the time-series was only for simple monitoring or querying, I would stay away from key-value databases like Prometheus or InfluxDB which have limited joins and limited analytics capabilities. A fully relational time-series database l
604.
by wenc 3y ago
You can Google it. Few people seem to know this and love to argue in circles about what H1bs make but all H1B salary data is public. You can even search by company and see how much your colleagues make for each job title (no names though) S
605.
by wenc 3y ago
Seems like someone has tried it: https://medium.com/@ahuarte/loading-parquet-in-postgresql-vi...
606.
by wenc 3y ago
I’ve found that there is a noticeable correlation between high HN karma and just poor behavior in general. Many of the least insightful and most nitpicky replies I’ve ever received were from people with karma > 10k. I wonder about why th
607.
by wenc 3y ago
Yes. My job involves pulling a ton of data off Redshift into Parquet files, and then working with them using DuckDB (sooo much faster — DuckDB is parallelized, vectorized and just plain fast on Parquet datasets)
608.
by wenc 3y ago
Spark is still pretty non performant. If the workload fits in memory and a single machine, DuckDb is so much more lightweight and faster.
609.
by wenc 3y ago
I mean there is Qt and wxWidgets but they look dated compared to Electron which supports the latest in web front end technologies.
610.
by wenc 3y ago
I don’t think I agree because it’s like saying accounting is just elementary level addition, subtraction and multiplication. But elementary school kids can’t do accounting. DL requires a certain degree of mathematical maturity to grasp.
611.
by wenc 3y ago
Why? It’s just a marketing term. Most predictive maintenance models are ML or statistical learning models and have a great track record. Some call it AI to sound hip (you can any ML model AI these days) but it’s probably just a standard mod
612.
by wenc 3y ago
It’s called predictive maintenance and it is the norm in many critical industries. Replacing equipment every fixed number of X years (no matter what the condition of the equipment is) is wasteful, costly disruptive and does not actually gua
613.
by wenc 3y ago
SpO2 (blood oxygen levels) is actually very useful if it's accurate. (Apple's implementation is not super accurate because SpO2 measurement from the wrist is fraught to start with -- it's usually measured from the finger -- a
614.
by wenc 3y ago
“The effective engineer” by Edmond Lau is a good book for the IC software engineer working in an engineering organization. The most influential content on engineering in my life is not in a book but a YouTube talk entitled How Complex Syste
615.
by wenc 3y ago
That is a very specific legal arrangement, which I understand is getting rarer and rarer these days. https://en.wikipedia.org/wiki/Jeonse
616.
by wenc 3y ago
On the contrary, ZN is taught is almost every undergrad control course — and not actually used industry. It’s a classic textbook method to demonstrate the concept in simulations but it has many downsides in practice (pushing the system to i
617.
by wenc 3y ago
If I had money and was looking for a high end toaster, I'd pay $200+ for a European toaster like SMEG. It'd be more for the aesthetics than the functionality. When you're paying that much money for a functional item, you&#x27
618.
by wenc 3y ago
Yes. Too many people who have only used the free ChatGPT 3.5 come away thinking, "if this is all there is, I've got nothing to worry about". It's not until they pay for ChatGPT 4.0 and use it for a while that they reali
619.
by wenc 3y ago
These are the stereotypes: (that mostly track my experiences living in different parts of the country) In the northeast, people are kind but not nice. They'll help you out when you have flat tire but may grumble under their breath that
620.
by wenc 3y ago
Don't let perfection be the enemy of the good. Today ChatGPT might only get you 80% of the way there, but you can still fill in the 20% and still come out ahead. For many of my tasks it's getting me 90% of the way there because th
621.
by wenc 3y ago
> I'm amazed at how little district heating is used in northern climates like Canada and the northern US. It is used on most if not all Canadian university campuses. Canadian university buildings are also far less sprawling because
622.
by wenc 3y ago
I used a bit of Mathematica but settled on Waterloo Maple because it had a cheaper academic license. Mathematica has stronger algorithms than Maple (like cylindrical algebraic decomposition), but for what I was using it for, they were both
623.
by wenc 3y ago
> It just doesn’t smell right that a team led by a guy who can’t manage to pronounce his own first name clearly could think through all the issues and have a coherent vision within a year of when LLMs really took off. Weird ad hominem. Y
624.
by wenc 3y ago
I don’t use Polars directly, but instead I use it as a materialization format in my DuckDB workflows. Duckdb.query(sql).pl() is much faster than duckdb.query(sql).df(). It’s zero copy to Polars and happens instantaneously while Pandas takes
625.
by wenc 3y ago
No protective case?
626.
by wenc 3y ago
Yep, going against the grain in graduate school is counterproductive unless there's a compelling reason. Many grad students forget that their main purpose is to generate research results and to publish papers that advance the field, no
627.
by wenc 3y ago
I use DuckDB for queries and Visidata for quick inspections. Between those two, I can work with not only CSVs, but also JSON and Parquet files (which are blazing fast -- CSVs are good for human readability and editability, but they're
628.
by wenc 3y ago
Looks nice but doesn’t fit in pocket. A less elegant but practical option for note taking in class or writing docs is a portable folding keyboard for iPhones, like this iClever Bluetooth Keyboard https://a.co/d/cibFqle
629.
by wenc 3y ago
Interesting. Curious if you were using GPT 4 or the free tier? I have a different experience. GPT4 has enhanced my reading by allowing me to correlate it with history, sociology etc. I don’t have the explicit objective of having my priors c
630.
by wenc 3y ago
Is that the Substack of the real Margaret Atwood?

More ›